|Name:||(11) Predictive Performance Tuning of OpenACC Accelerated Applications|
|Time:||Monday, June 23, 2014
05:32 pm - 05:39 pm
CCL - Congress Center Leipzig
|Speaker:||Saber Feki, KAUST|
|Abstract:||Graphics Processing Units (GPUs) are gradually becoming mainstream in supercomputing as their capabilities to significantly accelerate a large spectrum of scientific applications have been clearly identified and proven. Moreover, with the introduction of high level programming models such as OpenACC and OpenMP 4.0, these devices are becoming more accessible and practical to use by a larger scientific community. However, performance optimization of OpenACC accelerated applications usually requires an in-depth knowledge of the hardware and software specifications. We suggest a prediction based performance tuning mechanism to quickly tune OpenACC parameters for a given application to dynamically adapt to the current execution environment on a given machine. This approach is applied to a finite difference kernel to tune the OpenACC gang and vector clauses for mapping the compute kernels into the underlying GPU architecture. The performance results show a significant performance improvement against the default compiler parameters. Additionally, the predictive tuning is 18X to 52X faster than using a brute force search to explore the full parameter space.
Shahzeb Siddiqui & Saber Feki, KAUST